Electric Unicycle's BMS problem and solution

[Chriull]

Hi Rash - thanks for sharing this video!

Since i have a ninebot e+ myself, do you know which ninebot this is?

In the slow motion it seems, like the driver gave the ninebot an initial acceleration (bending his knee forward) before the ninebot stopped keeping the pedals straight. So could it be, that he overleaned the ninebot or was this just his last attempt to balance out the shut off of the ninebot?

...

Just to repeat one of the comments from youtube here: The wheel still spins (motor up an running) at 0:26 (after the fall) - so this was no shut-off but an overleaning?!

[Rotator]

But if the motor is big and the battery pack small, you can easily get into a situation where the controller now requests a very high current from the pack, to the point where the batteries are operating outside of the safe envelope, and voltage sags is magnified. In that case, voltage sags because the chemistry of the battery is being damaged. Not cutting the voltage to that cell is now bad, and can result in a cell fire. The worst part is that not all cells in a pack are identical, there always be one where the internal resistance is marginally higher. That call will experience the highest voltage sag of the pack, and at high currents be damaged marginally more, in turn increasing resistance and decreasing ability to safely provide current. Repeat a few times, and you now have a battery in the pack that will sag much more than all others, and operate more and more frequently outside of the safe envelope. Bad things will happen

Unless the main board knows how to limit the current draw for the whole pack, the individual mosfets for each battery are the only protection against over-current discharge. If the manufacturer relies on the BMS for over-current, you have a problem when shunting it

Based on Jason's empirical data, a good rule of thumb would be: if your battery Wh is at least half as your motor W, you should be safe making this change. If your motor W rating is higher than twice your battery Wh, you should think twice before doing it. If you have Jason's logger, you can instrument your specific EUC/battery pack and decide with 100% confidence if the shunting is safe or not

Bad things will happen if your EUC is powered off while running.

Bad things could happen if your shunted battery have an over-current discharge.

I ever thought that the protection mosfets should be used to power on a latched LED signal, beep signal or something like that, to lets the user know that protection has been triggered, and battery is not safe to be used and must to be replaced. But meanwhile, when the euc is running, the battery must supply every drop of power while it lasts, in order to keep the rider balance.

Hence, It is up to you to decide. Anyway, thank you for sharing this information, since can it help to choose a better decision.

 

[Chriull]

Hi, before all, let me congratulate you all for the amazing job, i already fell 2 or 3 times of my unicycle, at this moment i prefer ride on him catching fire than having him throw me away like a stone, -.-, i was almost  breaking my unicycle with a hammer when i saw this post, i´m not sure if its the same problem because, once it happens it doesnt turn on anymore, only when i get home and charge it, and since i have some daughts regarding this i decided to post them here,  hoping you guys can help me.

Sorry @hackerse7en I was quite sure that i already answered your post - but now i did not see it anywhere... Strange... Maybee I am just getting old

1- what is my model -.-'? Pictures Here

I have no idea - but maybe you find it on hobby16s blog. There he has gathered all specific shunting instructions for the wheels known to him. If you find it there, you can forget the rest of the post

2- Is the shunt in my picture correct, or should i shunt the 3rd one?

The 3 Mosfets could be right, but its only a guess and not for sure - see my comments/questions below.

The first problem is, that you shunt the wrong legs. You shall not, like painted in the picture connect the leftmost and the middle leg! According to hobby16s instructions you have to connect the rightmost leg with the middle leg! (Could veryfiy this with the data sheet of the P75NF75...)

3-I noticed over the mosfets it says "M19", instead of "Q1" or "T1", like i saw on some pictures, is that a problem?

Every producer can choose the names as he like.

If the three Mosfets have the same name, that would be a good hint. If they have different names, it just tells us nothing ;(

Do you feel/see somewhere under the plastic the other Mosfet(s) (X1 in hobby16s circuit diagramm)?

Are there more of these Mosfets just nearby (this group of three could be up to 6 Mosfets)

Can you make a picture which shows the PCB with the legs of this three Mosfets - maybe a more educated guess can be made once the PCB connections can be seen...

The main point is to indentify the Mosfets named T1 in hobby16 diagramm with certainty. Under no circumstance X1 should be shunted! If this happens to you, the accu pack will be overloaded while charging - and you do not want this happen!

From your fotos unfortionately this cannot be seen...

If you shunt one, two or three of the T1 Mosfets should not really matter, since they are in parallel. Maybe shunting all of them would be the best to divide the current. I do not have any idea, why hobby16 suggests from time to time just shunting 2 out of three? Could be that the third Mosfet could have been the one named X1?

Thanks a lot for the help, and have a safe ride
 

You, too!

[robca]

Bad things will happen if your EUC is powered off while running.

Bad things could happen if your shunted battery have an over-current discharge.

I ever thought that the protection mosfets should be used to power on a latched LED signal, beep signal or something like that, to lets the user know that protection has been triggered, and battery is not safe to be used and must to be replaced. But meanwhile, when the euc is running, the battery must supply every drop of power while it lasts, in order to keep the rider balance.

Hence, It is up to you to decide. Anyway, thank you for sharing this information, since can it help to choose a better decision.

 

Yes, it's a personal decision and it's not an easy one to make, granted. As you suggest, it might be possible to add a circuit so that a signal is provided (as an improvement on the simple shunting mod: instead of having the mosfet open the circuit, maybe there is a way to keep the 0.001 Ohm resistor and use it with a current sensing device to trigger an alarm. Basically shunt just the mosfet, not the mosfet and the 0.001 Ohm resistor, and add a way to detect when the voltage drop across that 0.001 Ohm resistor is too high). The risk of a simple shunt is that you will never know that a battery in the pack is getting into the danger zone, and keep riding until that battery explodes (and at that point you still will fall, since the pack will not provide any power at all).

I'm not good enough with analog electronics to suggest a clever way to improve on the shunt mod (I would build a pretty complex microprocessor-based circuit to measure the voltage drop and trigger the alarm), but I thought I'd flag this and see if some of the clever people here can make this mod even better... my suggestion is not to avoid the mod, it's simply to improve on the mod and not lose the over-current scenario

[Chriull]

Deleted - since robca posted a much more reasonable solution.

[robca]

I thought of a possible different option, if one doesn't mind soldering a few wires. It only costs ~$6 (plus waiting for Chinese eBay components to arrive)

I posted this in a separate thread http://forum.electricunicycle.org/topic/1024-monitor-individual-cell-health-for-your-battery-packlow-voltage-alarm/, not to derail the discussion here

[hackerse7en]

Sorry @hackerse7en I was quite sure that i already answered your post - but now i did not see it anywhere... Strange... Maybee I am just getting old

I have no idea - but maybe you find it on hobby16s blog. There he has gathered all specific shunting instructions for the wheels known to him. If you find it there, you can forget the rest of the post

The 3 Mosfets could be right, but its only a guess and not for sure - see my comments/questions below.

The first problem is, that you shunt the wrong legs. You shall not, like painted in the picture connect the leftmost and the middle leg! According to hobby16s instructions you have to connect the rightmost leg with the middle leg! (Could veryfiy this with the data sheet of the P75NF75...)

Every producer can choose the names as he like.

If the three Mosfets have the same name, that would be a good hint. If they have different names, it just tells us nothing ;(

Do you feel/see somewhere under the plastic the other Mosfet(s) (X1 in hobby16s circuit diagramm)?

Are there more of these Mosfets just nearby (this group of three could be up to 6 Mosfets)

Can you make a picture which shows the PCB with the legs of this three Mosfets - maybe a more educated guess can be made once the PCB connections can be seen...

The main point is to indentify the Mosfets named T1 in hobby16 diagramm with certainty. Under no circumstance X1 should be shunted! If this happens to you, the accu pack will be overloaded while charging - and you do not want this happen!

From your fotos unfortionately this cannot be seen...

If you shunt one, two or three of the T1 Mosfets should not really matter, since they are in parallel. Maybe shunting all of them would be the best to divide the current. I do not have any idea, why hobby16 suggests from time to time just shunting 2 out of three? Could be that the third Mosfet could have been the one named X1?

You, too!

is this a better view of the board? is it enough?
This are all the mosfets on the board, buth the numbers above them are really anoying m16, m17, m18 and m19

[Chriull]

is this a better view of the board? is it enough?This are all the mosfets on the board, buth the numbers above them are really anoying m16, m17, m18 and m19

 

If this are really all Mosfets that's good!

The photo is not enough - one cannot see how the connections at the PCB are.

All four Mosfets should have the middle leg connected.

Three out of this four have also the right and leftmost connected - these three you are searching for to shunt (T1).

The forth, which left and rightmost legs are not connected to the other mosfets left and rightmost legs, is not to be shunted (X1)! (Thats the one who allows you to charge your accu pack safely!)

If you have bad luck and they put 2 mosfets in parallel for T1 and 2 for X1, or the connections are not clearly to be seen, you have to look where leg number 1 (the leftmost) has a connection to a diode (labeled Dxx on the PCB), and from which mosfets leg 1 has a connection to a resistor (labeled Rxx  on the PCB...)

The other way would be to shunt B- to P- directly with a cable, like hobby16 showed for some examples.

[OhPardonMe]

Hi all. I bought one of the generic wheels off ebay not knowing about this issue. It has mohoo logos on the side. My second day of riding, I experienced a BMS cutout when leaning to accelerate while traveling uphill, causing a nice faceplant. 

I pulled apart the casing and it looks like my battery has 3 mosfets, seen below. These are the only mosfets I can distinguish through the battery casing. The center pin is not connected on any of the mosfets. 

Are these what I am looking for in order to shunt, and if so, which of these three should I be shunting? Thanks for the help.

[Chriull]

You can try to find your bms on hobby16 s blogsite - if its already on there, you'll find a detailed description. If not you can read my last comments regarding hackerse7en's bms shunting...

Edit: the middle pin is connected on all of them, or the bms would not work  either on the back of the pcb or some layer inbetween..

[Fred]

Hello guys,

I have a TG-F3 unicycle and things were going really well.

However, a couple of days ago, when I was going uphill though a bridge, I felt a scary unballance and I almost got a faceplant.

Luckyly I was able to recover the ballance on time and continue my way.

Since this day I´ve been thinking about options to reduce the risk of a BMS cut without getting the shunting risks.

First idea was if there is any replacement battery available to be purchased on the internet that doesn´t have the BMS.

This could fix every generic model (and even some old good ones) that has the problematic BMS circuit.

Second idea was to connect a generic 110v/220v switch (Picture) to each mosfet, in a way that in one position (ex: 110v) it would connect the 2 mosfet terminals, so it would be shunted, but in the other position (ex: 220v) it would not connect the terminals, so it wouldn't be shunted.

That way, when riding the EUC, we would "activate" the shunt, but when the EUC was inactive, we simply would "disable" the shunt.

Does that make any sense?

[esaj]

First idea was if there is any replacement battery available to be purchased on the internet that doesn´t have the BMS.

Bad idea, if there's no overcharge protection, the cells will catch fire or explode during charging. If you have a "smart" charger, that can handle the constant current and constant voltage-phases & balancing (needs a separate multi-pin connector from the battery), then it probably wouldn't be that much of a problem (unless the regenerative braking overcharges the cells, then it'll catch fire or explode under you )

Second idea was to connect a generic 110v/220v switch (Picture) to each mosfet, in a way that in one position (ex: 110v) it would connect the 2 mosfet terminals, so it would be shunted, but in the other position (ex: 220v) it would not connect the terminals, so it wouldn't be shunted.

That way, when riding the EUC, we would "activate" the shunt, but when the EUC was inactive, we simply would "disable" the shunt.

Does that make any sense?

The switch idea is good, but I think those switches you showed don't have high enough amperage rating (2A only). High powered wheels like MSuper can pull 1kW continuous and up to 3kW (or maybe even over) of power during peaks, and if the voltage is lower (say, 55V), that equates to around 18A continuous and 54A during peaks. The switch connectors need to be really sturdy.

[OhPardonMe]

You can try to find your bms on hobby16 s blogsite - if its already on there, you'll find a detailed description. If not you can read my last comments regarding hackerse7en's bms shunting...

Edit: the middle pin is connected on all of them, or the bms would not work  either on the back of the pcb or some layer inbetween..

I looked through the other thread and all posts in this thread and couldn't find one that was exactly the same. I am wondering, the main post references T1 and X1 mosfets or some variation thereof: Q1, Q2, etc. My question is, are the T1 and X1 required? Can there only be a T1 and no X1? or does this mean that one of the three mosfets in the picture I provided is the X1? Should I open up the battery casing fully to better verify what I am looking at or can anyone say with some confidence as to which of these I should be shunting just from the picture I provided?

[noisycarlos]

I looked through the other thread and all posts in this thread and couldn't find one that was exactly the same. I am wondering, the main post references T1 and X1 mosfets or some variation thereof: Q1, Q2, etc. My question is, are the T1 and X1 required? Can there only be a T1 and no X1? or does this mean that one of the three mosfets in the picture I provided is the X1? Should I open up the battery casing fully to better verify what I am looking at or can anyone say with some confidence as to which of these I should be shunting just from the picture I provided?

I also had a generic BMS with confusing markings (See my pictures on the previous page). When I asked @hobby16 how to know which mosfets to shunt he told me that the mosfets that are grouped together are usually the ones to shunt.

The mosfet that is usually by itself is the one coming from the charger (only one needed because there's less current) you must NOT shunt that one.

Here it is in his words:

To know what to shunt for sure, the rule is to look at the most paralleled mostfets and where it's connected to P- (power -) going to the mainboard : it's where most current flows. The mosfet NOT the shunt (usually alone) is the mosfet going to C- (charge -), less current flows there since it's connected to the charger delivering only 2A.

 

[Chriull]

X1 is the one which is never allowed to be shunted! You need it for save charging. Do not trust in any grouping or naming! Thats just a hint but nothing secure!

Grouping is a good hint, but if there are only these three mosfets on the board you have to look more detailed at the pcb.

[esaj]

This is kind of off-topic, but since this thread probably draws the electronics-specialists, does anyone know what I could do about the (probable) overcharge-problem? It looks like the BMSs of my new batteries shut off my wheel in downhill last night, as the batteries were fully charged and I started my journey downhill   Luckily the wheel just bounced some 10+ meters downhill and stopped in some bushes, I landed on my feet myself. 

Is there some sort of voltage divider/diode/transient-voltage-suppression diode/power-resistor burning excess energy off as heat-system I could work on the battery wires?   Btw, the BMS is the kind where the charging and discharging occurs over same wires, there are no separate "C-" -pads on the circuit boards.  I still need to test it again, but it seems that either the mainboard no longer limits (enough) the charge going back to batteries on regenerative braking or that the BMSs have "too low" overcharge protection, but I doubt it's the latter as it charged ok at 67.8V (According to Charge Doctor). And I most definitely don't want to do anything to the overcharge protections on the BMSs themselves, preferably I'd limit the voltage going back to the batteries.

I'd ask hobby16, but he's been hiding for the last month or so, hopefully he's just busy with work & his projects, and nothing bad has happened.

[hackerse7en]

well, since this is the only help i will ever get, hobby16 is out of the map. i decided to shunt p- to b-, and i tested the wheel, it seems that its working, however, when i started to charge it, the red light on the charger turns on, and after some time, is stays green, in other words, is charged, and then blinks red, and then green, and it keeps doing it. they blink until i remove it from the wall :/ . 
Here a pic of the board, before the shunt

[Chriull]

well, since this is the only help i will ever get, hobby16 is out of the map. i decided to shunt p- to b-, and i tested the wheel, it seems that its working, however, when i started to charge it, the red light on the charger turns on, and after some time, is stays green, in other words, is charged, and then blinks red, and then green, and it keeps doing it. they blink until i remove it from the wall :/ . 
Here a pic of the board, before the shunt

...

The battery was fully charged after you removed the charger? The manual says anything about the red/green blinking? (I assume it did not blink before your modification?)

You have a foto of your shunting p- and b-?

The foto you posted this time is good:

As it seems M18, M19 and M17 are T1 from hobby16s schematic and the ones to be shunted:

All three rightmost legs are connected (Source) and go to R71 & R69 (which should be R1 from hobby16s schematic - the current sense. R69 & R71 seem to lead to some kind of connector? (not clearly to be seen on the photo). This should be P- (the thick black wire to the mainboard).

M16 seems to be X1, which is not to be shunted:

here the source leads to R64/M15(r65) and then to the other side of the board. One of the two connections should lead to C-. Is this right?

Seem like they used a Transistor instead of the Diode D1 (can't read the Name of M15 on the foto - but seems to start with BC.. - the small connections there are not to be seen clearly, too - but that does not matter). Could also be that they have some different circuit for the overvoltage protection than the other BMS's (like hobby16s schematic). But as said, since nothing is modified there, it does not matter.

 

So you should have good chances by shunting M17-M19 to deactivate the overcurrent protection. 

But before you change something a foto of your work could maybe help to get your wheel working again without much effort.

B- should have a connection to all the middle legs of M16-M19 (there is a big "green" area, which leads to all of these middle legs - it should also connect to B-)

Could be that you did not shunt to B- but a wrong B.. (shunted one ore more cells away)? This could be the reason why you can drive (less than normal) and not really charge - the charger will give up sometimes if the batteries do not reach a specific voltage! (Just a guess - could be something completely different, too)

BTW: What is the "stain" below the legs of M19?

[SlowMo]

I just got a generic X3 clone and would also request for your expert advice. The unit arrived with only minimal battery charge. I charged it until full and have not used it until now. Upon reading this forum topic I decided to open the battery to check the mosfets and this is what I found. Is this battery safe to use? Thanks in advance.

[Chriull]

Seems like the leftmost is not to be shunted and the three rightmost are the ones to be shunted - but you should controll this with looking further under the plastic and look for the points i listed in my previous posts.

As long as alls the battery cells are in good condition, they have enough charge and one does drive "calmly" every battery pack-wheel combination is quite safe (regarding cut-offs). But with the over-current protection it could be (and is definitely with some modells) that you get unexpected cut-offs and have the possibility for a faceplant.

However - please read the previous posts and controll the points. Its your wheel, your money, your body - so your resposibility  I also could not quickly find a datasheet for this mosfet (for 99,999% they should be all pin compatible, but who knows...), and i dont have time for thorough checks/google - i am about to leave for a week...

[SlowMo]

 I was afraid just  by looking at the solder joints of this BMS that they seem to be soldered badly or have been damaged due to low voltage when the unit arrived. I will be careful and drive this unit slow with extra protection and  post a picture with a better view of the BMS for you to check when you come back from vacation. Thank you so much! 

I also placed an  order for the Charge Doctor before realizing  that Hobby16 is on vacation.

Edit: Got a reply from Fred. My charge doctor is already on the way! 

[PedalFaster]

Why do you keep repeating that a shunted battery has no undervoltage protection on monowheels ? That simply false. The mainboard always protects the battery from undervoltage, period.

I haven't seen the post this refers to, but there is a key difference between battery undervoltage and cell undervoltage.

A main board connected to the battery pack via two power wires alone can only measure the overall voltage of the pack.  If that is too low, the main board should definitely not be trying to run the wheel.  But with a high cell count pack, it is not necessarily sufficient information.

The potential issue with lithium cells concerns the voltage of each cell individually.  While manufacturers hopefully try to build packs out of cells that test having similar characteristics, inevitably they won't exactly match, especially once aged from use.  That means that a weaker cell can easily end up with a voltage notably below the "average" cell voltage suggested by the pack's series total voltage.

While a very conservative pack voltage monitor is better than nothing, only a circuit monitoring the voltage of each individual cell can fully protect against cell undervoltage (or, during charing, overvoltage).  On a typical wheel, the only thing that could do that job would be a BMS built into the battery pack; otherwise, it requires a mulitconductor wiring harness with a balance/monitoring tap in between each intermediate cell in the series string to grant that access to circuitry on the main board or external charger. 

[Chriull]

@hackerse7en 

Looks like ;( If you are experienced with soldering and have the right soldering iron it should be no problem to redo/repair the main soldering points. Especially the connecting points with the battery/mainboard/motor controll wires.

 

Imho there is no relation between low voltage and bad/damaged soldering points.

@PedalFaster

 

Shunting like hobby16 described does not have any influence on the low battery warning/alarm/protection. It only cancels the overcurrent (overdischarge) protection of the BMS. Also the overvoltage protection for charging is not touched in any way.

However, in newer BMS there could be a cell voltage monitoring be implemented, which could be triggered by the same MOSFETS that cut off in case of overcurrent.

Imho i would still shunt a BMS that cuts off, no matter for overcurrent or cell undervoltage. Better destroy the battery than doing a faceplant - i still feel my wrist from a normal, unspectacular slow motion fall in the beginning. I would not risk a broken wrist or something more serious, just to protect the battery or the mainboard. But as said - thats my humble opinion. On the other side, once i had my EUC shunted i would monitor the cell voltages while/before/after charging (or use a charger that monitors the cell voltages).

As far as i have read about Li Ion batteries, if you overdischarge them (normaly, like with an EUC) you just ruin or stress them. If you charge them (with the delivered charger for the EUC) after they are ruined/too much stressed they could get dangerous.

Edit: Multiquote did not work, so there are no quotes with my answers... ;(

[Tom]

I've read this whole thread and I'm still not 100% sure, could someone advise me:

I've got 2x 132wh batteries for my cheap 350w generic wheel, they are un-shunted and I'm keeping my speed below any beeps because I'm scared of BMS shutdown, but I want to remove that worry so I can push my EUC a bit faster. Would shunting my batteries mean that I can push the wheel harder with no chance of it turning off? If I shunt, does this mean that a damaged cell/battery will eventually catch fire, or are there other safety mechanisms in place that prevent it? Will there be any other changes to how the wheel performs that I will feel after shunting the battery?

Finally, is there any extra precautions I can take with a shunted battery so I know if it's no longer safe to charge/use (such as battery monitors/how to use?)?

 

Thanks!

[Tom]

Ok no replies  so I'm gonna go ahead and do it anyway. I felt 2 bulges on the battery, the one I opened in the image below and the other which feels more like a bar (heatsink?) Can someone instruct me as to what points to solder here? Do you need to see any more of the battery to be sure?

Also, I noticed the inner lining of the battery is paper/card! Is this safe?